Impact Methodologies

Distribution of Impacts

Forecasting Methods

Methods for discussing impacts such as accessibility or environmental impacts are discussed under the relevant impact sections. The following section identifies three techniques that can be used to disaggregate these impacts among socioeconomic groups or geographic areas.

Method 1. Spatially Based Analysis

This method compares the distribution of impacts among spatial units such as traffic analysis zones (TAZs) or census tracts, which can be classified by characteristic (low-income, predominantly minority, etc.). A general procedure is as follows:

1. Classify the spatial units according to the characteristic(s) across which the impacts are to be compared. For example, identify TAZs corresponding to census tracts with greater than X percent population in poverty or racial minority population.

2. Identify the magnitude of transportation project impacts for each spatial unit. For example, measure the change in accessibility, total emissions, or the concentration of emissions for each TAZ in the analysis area.

3. Compare the magnitude of impacts among the population groups of interest. For example, compare the average change in accessibility as a result of the regional transportation plan for TAZs with a higher racial minority population with the average change for TAZs with a lower minority population.

4. If appropriate, apply statistical tests to determine whether differences between alternatives and/or population groups are statistically significant.

Examples of this type of approach include:

• San Francisco Bay Area accessibility analysis (see case study): The accessibility of residents to employment was compared for "disadvantaged" and "non-disadvantaged" neighborhoods, under the Regional Transportation Plan versus the no-plan alternative. Accessibility was measured based on travel times from the regional travel model highway and transit networks and on forecast population and employment by TAZ. Statistical tests were applied to measure the significance of differences.

• Tren Urbano analysis (see case study): Accessibility is compared by mode (automobile, rail transit, and bus) and across five income groups to help analyze the impacts of the Tren Urbano rail transit project. TAZs are grouped according to average income level. For each group, the average accessibility of residents to employment is calculated based on transportation network travel times.

• The approach described above can be facilitated or enhanced through the use of a GIS. For example, emissions and noise contours can be developed from the locations and characteristics of roads and traffic, and these contours can be overlaid on spatial units such as census block groups or tracks. Then, impacts can be compared according to the characteristics of the spatial units. For an example, see the Waterloo case study.

• The Surface Transportation Equity Assessment Model (STEAM) is currently being modified by the Federal Highway Administration to assess benefits by zones. STEAM is a post-processor model that utilizes the output of four-step travel demand models and calculates various user benefits and externalities. Benefits and impacts will be assigned to TAZs or user-defined groups of TAZs to allow an analysis of the distribution of changes in impacts.

Method 2. Spatial Disaggregation

This approach is similar to Method 1, except that a GIS raster module is used to disaggregate socioeconomic data and impact data to grid cells. This allows impacts calculated for different types of spatial units to be more precisely overlaid on population data. For example, emissions from a transportation network can be assigned to the grid cells corresponding to the network, and then overlaid with population data that is assigned from the census tract level to each grid cell.

• Examples of the use of spatial disaggregation to identify the incidence of air quality impacts are provided in the SPARTACUS case study and the Envision Utah case study. The SPARTACUS case study also illustrates the distribution of noise impacts by population and compares the overall incidence of negative impacts across three socioeconomic groups.

Method 3. Microsimulation

Microsimulation travel modeling techniques forecast travel by modeling a set of actual or synthetic individuals or households that represent the population. (A "synthetic" sample is composed of a hypothetical set of people or households with characteristics that as a whole match the overall population.) Full microsimulation of a population is yet to be commonly implemented in practice due to computational requirements. A variant known as "sample enumeration" has been applied successfully in a number of areas, however. Sample enumeration relies on the modeling of behavior for a representative sample of the population.

The benefit of this modeling approach for analyzing the distribution of impacts is that travel patterns, and therefore the travel benefits of transportation improvements, can be tracked across any population characteristic that is included in the sample of persons modeled. Historically, this has been done by income level, since income is commonly used to predict travel behavior. The characteristics of the sample can also be broadened to include race or other characteristics.

Some examples of the microsimulation approach include:

• The STEP model, a sample enumeration-based travel forecasting approach. STEP has been applied in the San Francisco Bay Area in a number of studies, and has also been adopted for use in Los Angeles, Sacramento, Chicago, and Seattle. It has been used to analyze travel impacts of pricing scenarios by income group (EPA 1998), as well as for other purposes.

• A travel model is currently being developed for the City and County of San Francisco using a sample enumeration approach. The model is being developed specifically with the intent of tracking travel and benefits by race, income, and other characteristics (Cambridge Systematics, Inc., 2000).

The new generation of activity-based travel models (Engelke, 1997) also generally rely on sample enumeration, thus allowing benefits to be tracked by user characteristic. In an activity-based model, travel decisions become part of a broader activity scheduling process based on modeling the demand for activities rather than merely trips. An activity-based model was recently developed for the Portland, Oregon region and will be used in the future as their primary travel demand model (Bowman et al., 1998).

[TOP]